\begin{table}[!h]

\caption{Linear models on anticlerical violence\label{tab:lm_elec_compet}}
\centering
\begin{threeparttable}
\begin{tabular}[t]{lccc}
\toprule
  & (1) & (2) & (3)\\
\midrule
Employers' association & 0.260*** & 0.125* & 0.093\\
 & (0.057) & (0.056) & (0.091)\\
CNT union & 0.059 & -0.041 & 0.060\\
 & (0.044) & (0.050) & (0.044)\\
UGT union & 0.028 & 0.025 & 0.003\\
 & (0.026) & (0.025) & (0.023)\\
Employer $\times$ CNT union &  & 0.397** & \\
 &  & (0.133) & \\
Employer $\times$ UGT union &  &  & 0.272*\\
 &  &  & (0.113)\\
Population (log) & 0.118*** & 0.120*** & 0.118***\\
 & (0.016) & (0.016) & (0.015)\\
Electoral competition 1936 & -0.011 & 0.000 & -0.004\\
 & (0.028) & (0.029) & (0.028)\\
Days of Republican control (log) & 0.071*** & 0.071*** & 0.070***\\
 & (0.019) & (0.019) & (0.020)\\
Illiteracy 1930 & -0.413*** & -0.398*** & -0.399***\\
 & (0.090) & (0.088) & (0.091)\\
Number of clerics (log) & 0.261*** & 0.258*** & 0.258***\\
 & (0.029) & (0.029) & (0.028)\\
\midrule
$n$ & 3948 & 3948 & 3948\\
$R^2$ & 0.44 & 0.44 & 0.44\\
Adj. $R^2$ & 0.43 & 0.44 & 0.43\\
Province FE & Yes & Yes & Yes\\
\bottomrule
\end{tabular}
\begin{tablenotes}[para]
\item \textit{Note: } 
\item + p $<$ 0.1, * p $<$ 0.05, ** p $<$ 0.01, *** p $<$ 0.001. Province FE not shown, clustered SE (province). Only municipalities that were at least one day under Republican control.
\end{tablenotes}
\end{threeparttable}
\end{table}
